Unveiling Agricultural Soil Runoff: Remote Sensing and Ensemble Deep Learning Models to Evaluate Impact of Climate on Water Quality and Human Health

Water quality monitoring is a critical aspect of environmental management, as it directly impacts public health, ecosystem sustainability, and resource management. Traditional approaches to predicting water quality often rely on individual machine learning models, which may not fully capture the complex temporal and non-temporal relationships inherent in environmental data. This study presents a novel Ensemble Deep Learning Model (EDL) that combines Recurrent Neural Networks (RNNs) and Random Forests to predict the impact of climate on water quality parameters with high accuracy. The proposed model leverages the strengths of RNNs in capturing temporal dependencies and the robustness of Random Forests in handling non-linear relationships and feature importance. The model was evaluated on a comprehensive dataset containing various water quality indicators such as pH, turbidity, and dissolved oxygen which changes based on climatic conditions. The results demonstrate the model’s superior performance, achieving an accuracy of 97.81%, significantly outperforming individual models. The model was executed in PyCharm on a Windows system with an Intel® Core™ Ultra 9 Processor 185H, ensuring efficient processing and high computational performance. This study highlights the potential of ensemble models in environmental monitoring and provides a reliable approach for predicting critical water quality parameters that impact human health and ecosystem sustainability.